The proposed project is a long-term evaluation of an ongoing community- based drug abuse prevention program Midwestern Prevention Project) that has emphasized intervention with young during early adolescence,and experimental test of additional program components as these youth move into late adolescence. Analyzed through three year follow-up, the program has produced significant reductions in monthly and weekly use of tobacco, alcohol, marijuana; by three year follow-up, effects on monthly cocaine use and lifetime amphetamine use are also significant. Effects are significant regardless of gender, race, urbanicity, SES, grade, or baseline levels of drug use. The proposed project will follow these subjects through late adolescence and early adulthood in two studies: Study 1 continues the tracking of an existing panel of adolescents in Kansas City, and Indianapolis, half of whom previously received a comprehensive community- based intervention that includes school, parent, community organization, health policy change, and mass media program components (N=2,000; N=27 high schools). Study 2 is a longitudinal evaluation of new panels of adolescents who will enter the 9th grade high school transition year or the 12th grade pre-transition year in 1990-91, all of whom received the community program during early adolescence, half of whom will be randomly assigned by school to a 9th or 12th grade booster program, (N=4,000; N=32 high schools); The 9th grade panel from Indianapolis will be used as an external comparison (N=1,000; N=17 high schools). Study 1 will produce longitudinal data on the maintenance of community program effects from ages 11 through 21, i.e., through legal drinking age and early adulthood, and replication in two cities. Study 2 will evaluate short- and long-term effects of strategic" program boosters that are timed to coincide with the transition to high school (ninth grade) and preparation for transition to college or worksite (12th grade). Measures include annual self-report, expired air, school archival records on drug-related infractions, and community archival data on drug use-related morbidity and mortality rates for the age group 15-23. Regression, casual modeling, and time series analysis will be used to evaluate program effects and monitor community trends. The proposed project continues the work of the current project by using the same subject pool, data collection procedures, theoretical program content, and continued evaluation of the effects of early preventive intervention and continued community support. The results of the study should yield important information about: (1) drug use etiology in a large population representative sample of individuals across a 10-year period of the lifespan that encompasses three drug use risk periods; (2) maintenance of effects of school and community programming delivered during early adolescence; and (3) the potential for program boosters to enhance and maintain program effects through early adulthood.